Pan-zoom Motion Capture in Wide Scenes using Panoramic Background

Masanobu Yamamoto

2022

Abstract

Measuring a subject three-dimensionally from multiple cameras, the measurable area is a common field of view from cameras. When the subject goes out of the field of view, the cameras must follow the subject. In this research, the viewpoint of the camera keeps to be fixed and the pan and zoom functions are used to operate the cameras so that the subject body could be always shot near the center of the image. The problem is camera calibration. Our approach is to use a panoramic image. Each camera pans in advance to take a background image and create a panoramic image of the background. Then, the background image around the subject body is collated with the panoramic image, the pan rotation angle and the zoom ratio are obtained from the matching position, and the camera is calibrated. The body motion is captured from the multi-view motion image using the camera parameters obtained in this way. Since the viewpoint of the camera is fixed, the shooting range is not so wide, but it is still possible to capture an athlete’s floor exercise in the gymnasium.

Download


Paper Citation


in Harvard Style

Yamamoto M. (2022). Pan-zoom Motion Capture in Wide Scenes using Panoramic Background. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 1: GRAPP; ISBN 978-989-758-555-5, SciTePress, pages 163-171. DOI: 10.5220/0010815600003124


in Bibtex Style

@conference{grapp22,
author={Masanobu Yamamoto},
title={Pan-zoom Motion Capture in Wide Scenes using Panoramic Background},
booktitle={Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 1: GRAPP},
year={2022},
pages={163-171},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010815600003124},
isbn={978-989-758-555-5},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 1: GRAPP
TI - Pan-zoom Motion Capture in Wide Scenes using Panoramic Background
SN - 978-989-758-555-5
AU - Yamamoto M.
PY - 2022
SP - 163
EP - 171
DO - 10.5220/0010815600003124
PB - SciTePress